Source: Nature
ArXiv is tightening submission limits to stem AI-generated and low-quality papers flooding the platform. The move reflects a core tension: the same large language models that speed up research also enable frictionless paper generation at scale, forcing gatekeepers to weigh openness against quality control. Similar pressure may push other preprint servers—bioRxiv, SSRN—toward comparable throttling, potentially fragmenting where early-stage research gets published based on submission discipline rather than scientific merit alone.